Augmenting input at time t based on output at time t-1

Element-Research / rnn

Recurrent Neural Network library for Torch7's nn

BSD 3-Clause "New" or "Revised" License

939 stars 313 forks source link

Augmenting input at time t based on output at time t-1 #367

Open suryabhupa opened 7 years ago

suryabhupa commented 7 years ago

Normally, when decoding with an LSTM, the output at time t-1 is used as the input at time t. Instead, I want the input at time t to be [output{t-1} f(output{t-1})], where f is feedforward neural network. Does anyone know how to cleanly incorporate this into the code?

squidszyd commented 7 years ago

same question